ACASVA is a joint research venture bringing together interdisciplinary scientific and engineering expertise at the Centre for Vision, Speech and Signal Processing (CVSSP) and the Department of Psychology at Queen Mary University of London, and the School of Computing Sciences (CMP) at the University of East Anglia. This project will address the challenging problem of autonomous cognition at the interface of vision and language.The main aim of this work, in correlation with current ACASVA (Adaptive Cognition for Automated Sports Video Annotation) project, is to build such a system focusing specifically on court-based games, such as Tennis and Badminton. While the main implemented system has been trained for a specific game like Tennis, the goal of our work is to adapt the players action recognition module for a new domain such as Badminton or even Table-Tennis and other court based games in the future without an explicit system redesign. Actions are derived using the HOG3D feature extraction method and for transfer learning we used a method based on feature re-weighting and a novel method based on feature translation and scaling.
CDL or charting the digital lifespan combins social, technical, design, and cultural expertise, considering what it means for individuals to ‘live out’ digital lives across the complete human lifespan. I had contributed in statistical pattern classification and multi-view learning, with applications in language and image processing. I collaborated with Stuart James in the context of large-scale social media classification via the fusion of text and image.
CityPulse provides innovative smart city applications by adopting an integrated approach to the Internet of Things and the Internet of People. The project will facilitate the creation and provision of reliable real-time smart city applications by bringing together the two disciplines of knowledge-based computing and reliability testing.
(note: Click the MAP for a live demo of mapped Tweets from the London city.)